I am attempting to run a complex transient simulation. I have a design which relies completely on natural convection for cooling.
To simplify and allow the transient simulation to be run in a reasonable amount of time I am trying to set up a conduction only model. I ran a model with convection for a steady state case. I am attempting to extract the film coefficients from the outside surfaces to apply to the conduction only model (applied as outer wall boundary conditions).
Unfortunately the number I am getting from the surface parameter extraction seem far off from what they should be. I am attempting to use the average film coefficient number. The extracted average numbers range from 1.2-2.8 W/m^2-K. I know from hand calcs that these numbers are low and should be in the range of 9-20 W/m^2-K.
When I run the conduction only model with the film coefficient values from the steady state model the temps are about 100C hotter than they should be. When I increase the film coefficient to values closer to what they should be, the numbers become more reasonable.
Any ideas on why the extracted surface parameters are so far off? Has anyone else had problems with film coefficient values?